Polsinelli attended the February 18, 2020 Compensation Roundtable jointly hosted by OFCCP and the National Industry Liaison Group (NILG).  The event provided valuable insight into the thought processes of senior OFCCP officials as they fielded questions from federal contractor compliance professionals.  The frank discussion during the roundtable highlighted contractors’ ongoing concerns regarding OFCCP’s approach to the evaluation of the compensation of small groups of specialized employees and smaller AAP groups.

OFCCP’s preferred method of evaluating compensation and other personnel practices is the regression analysis, a statistical tool that seeks to control for various non-discriminatory factors and identify statistically-significant disparities that may be based on race, gender, or other protected characteristics.  Regression analyses can be effective for evaluating large groups of employees performing the same job, such as entry-level or line employees at larger organizations, but is an ineffective tool for addressing employees in higher-level, specialized positions, which may have only one or a small handful of employees or smaller AAP establishment groups consisting of relatively few employees.  Two common contractor complaints regarding OFCCP’s statistical analysis were highlighted in the Roundtable: (1) OFCCP’s combination of groups of employees across job titles, salary grades, job functions, or other relevant groupings into pay analysis groups (PAGs), and (2) OFCCP’s failure to properly consider and control for non-discriminatory factors relied on by the contractor in setting compensation. 

Both of these concerns stem from OFCCP’s reliance on the regression tool, which OFCCP contends requires a large number of data points to provide reliable results.  It is OFCCP’s position that valid PAGs must contain at least 30 employees.  For specialized or higher-level positions, this rule can result in PAGs containing numerous different positions having different job functions in different departments.  The contractor advocates at the Roundtable urged OFCCP to drop its insistence on 30-member PAGs in every case and instead analyze more focused “similarly-situated employee groups” (SSEGs) that do not consolidate employees in dissimilar positions.

Although OFCCP indicated that it will control for differentiating factors among PAG members, two other statistical principles relied upon by the agency hinder its ability to effectively do so.  First, OFCCP requires that for each control factor, the PAG contain at least 10 employees (i.e., to control for 5 different factors requires a PAG of at least 50 employees).   This creates a Catch-22 in which the creation of a PAG requires consolidating dissimilar positions, but including additional controls necessary to account for the dissimilarity requires adding still more employees in other positions to the PAG.  The second rule applied by OFCCP is that any variable must have at least 5 observations or data points or OFCCP will combine that variable with the next closest variable.  For example, if an employer’s performance evaluation system rates employees into three categories – Needs Improvement, Meets Expectations, and Exceeds Expectations – and only four employees in the PAG are rated as “Needs Improvement,” those four employees will be combined with the employees in the “Meets Expectations” category, eliminating consideration of a factor that is likely significant to the employer’s decisions.  Although the OFCCP officials at the Roundtable expressed some openness to considering small group analysis techniques, they showed little willingness to move away from the agency’s current approach to these principles.

In addition to recommending that OFCCP adopt a more flexible approach and reduce its rigid adherence to these statistical principles, the contractor advocates at the Roundtable also criticized the agency for failing to consider two factors commonly relied upon by employers in setting compensation: job grades and market rate.  The contractor representatives noted that an employer’s salary or job grades represent the employer’s holistic business judgment about the relative value of each position within its organization. OFCCP, on the other hand, contended that if the agency has data regarding the underlying factors that inform the assignment of employees to grades (i.e., job responsibilities, seniority, etc…) then the agency should more reliably analyze those underlying factors.  As the contractor representatives noted, however, OFCCP’s independent assessment of those factors to the exclusion of the grades actually used by the employer may be inconsistent with OFCCP Directive 2018-05’s emphasis on conducing compensation evaluations using the same factors relied upon by the employer.

The Compensation Roundtable provided a forum for a frank discussion of OFCCP’s compensation evaluation practices.  Although there is a wide gulf between the views of OFCCP officials and contractor advocates regarding OFCCP’s analytical techniques, the agency’s consideration of contractors’ concerns is a step in the right direction.  OFCCP’s apparent reluctance to move away from some of its current practices identified as contractor concerns highlights the need for contractors to rigorously self-audit their compensation practices to identify potential agency concerns and address them prior to being selected for a compliance evaluation.